Growth and development

Larval SL increased from 4.4 to 4.6 mm by 5 dph and then remained more or less unchanged, whereas DM steadily decreased from 68.7 to 27.9 to 28.9 μg from the start until the end of the experiment (Fig. 3). Yolk-sac area also decreased throughout the experiment (from 1.39 to 0.015–0.027 mm2) and yolk absorption rates, as calculated from exponential regression lines, were –0.451, –0.442 and –0.441 d–1 in the NO, NA and CW treatments, respectively.

Myotome heights were between 0.21 and 0.24 mm and changed little throughout the experiment. For each of these 4 measures (SL, DM, YSA and MH), the vast majority of daily treatment comparisons was not significantly different except in 4 instances: (1) SL at 10 dph (NA > CW), (2) DM at 10 dph (CW > NO and NA), (3) YSA at 4 dph (NA > NO and CW) and (4) MH at 9 dph (CW > NO and NA).

4. Discussion

The present study clearly demonstrated precocious and more intense exogenous feeding by the larvae of a marine fish species in the presence of autotrophic phytoplankton (Nannochloropsis sp.) and heterotrophic dinoflagellates (Oxyrrhis marina). Young larvae in our CW, NA and NO treatments started first feeding at 6, 4 and 3 dph, had 11, 42 and 60%

yolk reserves remaining on that day, and reached a peak of 68, 75 and 95% first feeding on 7, 6 and 5 dph, respectively. The presence of autotrophic phytoplankton and heterotrophic protists, therefore, corresponded to a substantial increase in first-feeding success and expansion of the WOO for first feeding. Despite these benefits, our experiment also revealed that the nutritional value of these protists was not great enough to alter the (macroscopic) growth characteristics of yolk-sac larvae relative to conspecifics maintained in pure salt water.

The results of the present study agree well with those of previous research reporting beneficial effects of rearing marine fish larvae in the presence of microalgae including (1) earlier first feeding (Kentouri 1985, Naas et al. 1992, Maurizi 2000), (2) reduction of metabolic stress and prolonged survival of unfed larvae (reviewed in Muller-Feuga et al. 2003) and (3) dramatic improvement of early survival in marine fish species considered problematic to rear (Naas et al. 1992, Reitan et al. 1997). After first feeding, the presence of algae may also (4) increase rates of food consumption and/or growth (Øie et al. 1997, Bengtson et al. 1999, van der Meeren et al. 2007). Previous work by van der Meeren et al. (2007) indicated increased gut filling at 3 dph (7 to 8°C) in cod larvae exposed to microalgae, which agrees with the results

obtained in the present study. However, the present study indicates that this improved gut filling is sustained throughout the mixed feeding stage until PNR is reached. The mechanism responsible for our GFI findings is not known but could involve either intrinsic and/or extrinsic (environmental) factors. For example, different combinations of light intensity and algal concentration may enhance visual contrast, improving prey capture in larval fish (Boehlert & Morgan 1985, Naas et al. 1992). It is important to note that the differences in feeding intensity observed among treatments in the present study cannot be attributed to the small differences in light intensity (175 lux) experienced by larvae in CW compared to NA and NO groups. No significant differences in first feeding by cod larvae were observed over a larger range in light intensities (~50 to 690 lux) in the presence or absence of microalgae (van der Meeren et al. 2007). The difference in feeding intensity between our NA and NO groups (which were at the same light and algal densities) supports the overriding effect of algal characteristics over light intensity in this particular study.

This stimulation of early feeding could be linked to the drinking phase, an important preliminary phase that maintains osmotic balance in the developing larvae as well as instigating absorption of dissolved organic material and ingestion of some small particulate matter (Muller-Feuga et al. 2003). During the drinking phase, cod have been reported to ingest dinoflagellates either passively (via drinking) or actively (by filter feeding) (Ellertsen et al. 1980, van der Meeren 1991), and evidence of some nutritional benefit from imbibing algae has been observed, including changes in phospholipid composition and triglycerol content in first-feeding larval cod attributed to the presence of microalgae (Isochrysis sp.) (van der Meeren et al. 2007). After first feeding, cod larvae have been shown to feed on small protozoans but this appears to be more energetically costly compared to feeding on larger copepod nauplii (Hunt von Herbing et al. 2001). Thus, our finding of no direct (macroscopic) nutritional benefit, such as decreased yolk/energy utilisation or improvement in larval condition indices (dry mass- and myotome heightat- age) among larvae in any feeding treatment, may not be surprising. Regardless of a lack of direct growth benefits, stimulation of first feeding by microalgae and heterotrophic protists may be particularly important for the survival of cod and perhaps in other marine fish species that have a relatively brief WOO.

Larvae in the CW treatment started to first-feed (reached FI50) at 60 degree-days post-hatch (ddph = °C·dph), and at 63 ddph, unfed cod larvae exhibited morphological changes in gut villi detrimental to digestive and absorp tive capacity (Kjørsvik et al. 1991). Furthermore, the earlier feeding and increased GFI when feeding would have likely caused NO and NA larvae to exhibit more rapid growth compared to CW larvae after first feeding, but this was not tested in the present study.

Although direct nutritional benefits of algae and heterotrophic protists to larvae were lacking in the present study, their presence appeared to ‘prime’ biochemical and/or developmental systems that, in turn, promoted first feeding. During early development, larval fish cannot synthesise phospholipids (Bell et al. 2003) and must acquire these through their diet. Algae have been demonstrated to stimulate the production of digestive enzymes such as amylase and trypsin, characteristically the first enzymes to be recorded in early marine fish ontogeny (Cahu et al. 1998). It was hypothesized by Cahu et al. (1998) and other authors (Støttrup 1994) that the large amounts of free amino acids contained in algae could be responsible for

the observed stimulation in the production of trypsin. In the wild, free amino acids are often at concentrations <10–7 M (Braven et al. 1984) but can occur at higher concentrations in areas of high phytoplankton production (Williams & Poulet 1986). Therefore, intuitively, one could link these dissolved free amino acids to the stimulation of feeding of fish larvae in the wild.

This seems reasonable since concentrations of algae and protists used in the present laboratory study were of the same magnitude as those occurring in situ (e.g. Arndt 1991, Tamigneaux et al. 1997, Hansen & Jensen 2000) at the locations and during times of the year when first-feeding cod larvae would be expected to occur. Naturally, other benefits from microalgae and protozooplankton such as improvement in larval microbial gut flora (Skjermo & Vadstein 1993) should not be dismissed.

The results of the present study highlight the need to revisit the importance of autotrophic phytoplankton and heterotrophic protists in the early survival and growth of marine fish larvae. Previous studies indicated that the larvae of some marine fish species routinely ingest phytoplankton (e.g. Northern anchovy Engraulis mordax feeding on dinoflagellates) which appeared to offer some nutritional benefit (e.g. Scura & Jerde 1977) and was thought to be important for early survival and recruitment success (Lasker 1975). Our results clearly indicated that cod obtained no nutritional (growth) benefit from ingesting algae and heterotrophic protists (protozooplankton) prior to first feeding. However, the presence of adequate amounts of protozooplankton may be critical to the survival and growth of cod by acting to prime first-feeding capabilities of this species, altering the window of opportunity for successful firstfeed and, in turn, influencing the match-mismatch dynamics between first-feeding cod and their macrozooplankton prey.

5. Figures

Figure IV-1 Mean feeding incidence (FI) in yolk-sac larvae versus larval age in the presence (NA) or absence (CW) of algae (Nannochloropsis sp.) or in the presence of algae and Oxyrrhis marina (NO). Significant differences in FI were found among treatments at ages 3 to 6 d post-hatch. #Days with significant differences among treatment values. Different letters (a, b) denote significant differences (ANOVA, Tukey post hoc test, p ≤ 0.05, n = 3 replicates treatment–1). The ‘window of opportunity’ and magnitude of FI are also indicated (shaded areas) calculated from regression equations (Eq. 2 in ‘Data analysis’ under ‘Materials and methods’; see Table 1). Error bars indicate SD (n = 3).

Figure IV-2 (A–C) Example of gut fullness index (GFI) scores in first-feeding larvae, as typed either ‘1’ (<6 intact copepods and/or remnants in the gut), ‘2’ (>6 clearly distinguished copepods in the gut but a gut that was not distended) or ‘3’ (fully distended gut, packed with prey). Photos taken at 12× magnification. (D) Mean GFI scores in yolk-sac larvae versus age for larvae reared in the presence (NA) or absence (CW) of algae (Nannochloropsis sp.) or in the presence of algae and Oxyrrhis marina (NO). On a given sampling day, significant differences in mean GFI are indicated with different letters (ANOVA, Tukey post hoc test, p

< 0.05). No significant differences among treatments were found at 3, 8 and 11 d post-hatch.

Error bars indicate SD (n = 3 tanks).

Figure IV-3 Mean standard length, freeze-dried mass, yolk-sac area and myotome height versus age for yolk-sac larval cod reared in the presence (NA) or absence (CW) of algae (Nannochloropsis sp.) or in the presence of algae and Oxyrrhis marina (NO). On a given sampling day, different letters denote significant differences (ANOVA, Tukey post hoc test, p

< 0.05, n = 3 replicate tanks treatment–1). Error bars indicate SD (n = 3).

6. Tables

Table IV-1 Summary information for feeding incidence (FI) by yolk-sac cod larvae reared at 10°C in only seawater (clear water, CW) and in seawater including Nannochloropsis sp. (NA) and both Nannochloropsis sp. and Oxyrrhis marina (NO). Parameter estimates and statistics for Eq. (2) (see ‘Data analysis’ under ‘Materials and methods’) are provided: maximum mean feeding incidence (FIMAX), slope (b) and age of larvae at FIMAX. Calculated values include the age of larvae at 50% feeding incidence (t50), the point of no return (PNR50), window of opportunity (WOO) and the relative magnitude of feeding (see Fig. 1). FI was expressed as the mean of 3 replicate tanks. dph: days post-hatch

Literature

(Arndt, 1991; Bell et al., 2003; Bengtson et al., 1999; Blaxter and Hempel, 1963; Boehlert and Morgan, 1985; Braven et al., 1984; Cahu et al., 1998; Cushing, 1975; Diaz et al., 1998;

Ellertsen et al., 1980; Fukami et al., 1999; Hansen and Jensen, 2000; Hjort, 1914b; Houde, 2008; Howell, 1979; Hunt von Herbing et al., 2001; Kentouri, 1985; Kjørsvik et al., 1991;

Lasker, 1975; Maurizi, 2000; Muller-Feuga et al., 2003; Munk, 1997; Naas et al., 1992; Øie et al., 1997; Peck and Holste, 2006; Pepin and Penney, 2000; Pepin and Dower, 2007; Ptacnik, 2003; Reitan et al., 1997; Scura and Jerde, 1977; Skiftesvik et al., 2003; Skjermo and Vadstein, 1993; Sommer et al., 2002; Støttrup, 1994; Tamigneaux et al., 1997; van der Meeren et al., 2007; Williams and Poulet, 1986; Yufera and Darias, 2007)

Discussion

The fragile early life stages of marine fish larvae have attracted much scientific attention ever since their relevance for recruitment success and fisheries had been realized. Cod (Gadus morhua) has not only become a model species for biological oceanography, but this species also justifies its use in historic and modern fisheries research by its unmatched economic importance. A wealth of knowledge on the biology of the species was accumulated in the last centuries and still new discoveries are to be made. The present work is a contribution to the biological understanding on one of the minor sized populations of cod, located in the eastern part of the Baltic Sea.

This more-or-less genetically distinct population is currently considered one of the more productive fisheries stocks and has undergone severe fluctuations in the last decades. The unique hydrographic conditions of its brackish habitat have pushed the stock, and the related fisheries industry, to the borders of its economic extinction and the same environmental factors brought the population back to life. This work was carried out in the context of two research projects: UNCOVER (“Understanding the mechanisms of stock recovery”, EU-FP 6, Contract number 022717) and RESTOCK (“The production of Baltic cod larvae for restocking in the eastern Baltic”, Financial Instrument for Fisheries Guidance, FIFG). The former sought to understand the mechanisms of stock recovery and the latter sought to restock the eastern Baltic cod population by the release of newly hatched yolk sac larvae. Both projects had a common sense on the importance of in-depth understanding of the effects of key environmental factors on cod larval growth and condition. The mandate to conduct this work was therefore very clear: controlled laboratory experiments with eastern Baltic cod larvae are needed.

The experiments were conducted on the island of Bornholm, the pearl of the Baltic, at the western boundary of the population’s distribution range. Two years of intense laboratory experimental seasons were needed to gather enough data to present this thesis. The author has spent 13 months on the island, during the summer-time spawning season, and has intensely worked with the broodstock and egg and larval material provided by Bornholm’s Lakseklækkeri, Nexø, Denmark. Sample processing and data analysis was conducted at the Institute of Hydrobiology and Fisheries Science, University Hamburg and at DTU Aqua, Copenhagen.

The experiments were grouped in two thematic areas. The first block of experiments, mainly conducted in the year 2007, investigated the effects of food and temperature on larval cod vital rates (growth and condition) and the resulting implications for fisheries science. The second group of experiments in the year 2008 investigated the effects of Alizarin Complexone marking, a method of pivotal importance for the assessment of stocking success, and innovative “green water” techniques to enhance larval fitness in aquaculture. The results of these experiments are collated in this thesis and can be found as peer-reviewed scientific publications (three published manuscripts; one manuscript draft).